Abstract
Nerve growth factor (βNGF) is a protein supporting sympathetic and sensory innervation in the peripheral tissues as well as cholinergic innervation in the brain. A DNA probe derived from a genomic clone coding for chicken NGF was used to study NGF mRNA levels during development. NGF mRNA was detected in the chicken embryo as early as day 3·5 of incubation. The level of NGF mRNA in total embryo increased four-fold until day 8, remained high until day 12, and subsequently decreased. No corresponding peak in NGF mRNA expression was found in heart and brain measured separately. Instead these organs showed increased NGF mRNA levels after hatching. The highest levels of NGF mRNA in the day-8 embryo were found in skin and eye (in particular cornea, but also iris, sclera-choroid and neural retina) suggesting a correlation between sensory innervation and this early peak of NGF expression.
Introduction
Nerve growth factor supports sympathetic and sensory innervation of peripheral tissues (Levi-Montal-cini & Angeletti, 1968; Thoenen & Barde, 1980). Recently NGF has also been found in the mammalian brain (Korsching et al. 1985; Whittemore et al. 1986; Shelton & Reichardt, 1986; Goedert et al. 1986) and attributed a trophic function for cholinergic neurones situated in the basal forebrain (Gnahn et al. 1983; Korsching et al. 1985; Large et al. 1986). The recent isolation of the gene for chicken βNGF (Ebendal et al. 1986; Meier et al. 1986; Wion et al. 1986) has made it possible to document the presence of NGF mRNA also in the avian brain (Ebendal et al. 1986; Wion et al. 1986; Goedert, 1986). However, the developmental appearance and function of NGF in avian embryos are still largely obscure, despite the fact that the classical biological assay for NGF (Cohen et al. 1954) is based on stimulation of neurite formation in embryonic chicken ganglia. In this report, we have used a DNA probe for chicken NGF to address the question of when and where NGF mRNA is synthesized in the chicken embryo.
Materials and methods
RNA blots
Total RNA was prepared from entire embryos (embryonic day (E) 3·5-18), and organs from embryos (E8-18), or posthatch chicks (day Pl and P7), as previously described (Ebendal et al. 1986). Poly(A)+ RNA was purified by oligo(dT)-cellulose chromatography (Aviv & Leder, 1972) and used for RNA blots.
10 μg of poly(A)+ RNA from each sample were electrophoresed in 1 % agarose gels containing 0·7 % formal-dehyde, transferred to nitrocellulose filters and hybridized to a 900 bp Pstl fragment encoding the mature chicken NGF protein (Ebendal et al. 1986). The probe was labelled by nick translation using α-32P-dCTP to a specific activity of around 5x108ctsmin-1μg-1. Filters were washed at high stringency (15mM-NaCl, l-5mM-sodium citrate, pH 7·0, 0·1% SDS, 54°C) and exposed to Kodak XAR-5 films at —80°C using Du Pont intensifying screens. Filters were then boiled for 10 min in 1 % glycerol and rehybridized with a probe for mouse α-actin as a control for the amount of RNA in each lane.
Autoradiograms were scanned using a Shimadzu CS-390 densitometer and the linearity was checked by scanning of serial dilutions of RNA obtained from COS cells transfected with a eukaryotic expression vector containing the chicken NGF gene (Ebendal & Persson, 1987; Hallbôok et al. 1987).
Results
Developmental changes in NGF mRNA levels
A DNA fragment from a genomic chicken NGF clone (Ebendal et al. 1986) was used as a probe to measure the levels of NGF mRNA in the chicken embryo. Blots of RNA prepared from total embryos at different ages showed the presence of NGF mRNA already at E3-5 (stage 20; Hamburger & Hamilton, 1951), the earliest stage examined (Fig. 1). The size of the embryonic NGF transcript was the same as in the adult (1-3 kb; Ebendal et al. 1986). In addition, weak hybridization was seen to a 5 kb RNA species that may represent a NGF RNA precursor or cross-hybridization of the probe to ribosomal RNA (Fig. 1). The level of NGF mRNA in total embryo increased slightly at E4·5 (stage 25), showed peak levels at E8 and E12, and was markedly lower during late embryonic development (E18) (Figs 1, 2) .
For comparison, levels of NGF mRNA were studied also in heart and brain from E8 until one week after hatch. These two organs were previously found to contain higher NGF mRNA levels than most organs in the adult chicken (Ebendal et al. 1986). Low levels of NGF mRNA were detected in the embryonic heart (E8, E12 and E18), whereas increasing levels were found after hatch (Fig. 2). Similarly, brain showed low levels of NGF mRNA throughout the embryonic period. At E8, the level of NGF mRNA in the entire embryo was around ten times higher than the levels found at the same developmental age in the brain and heart (Fig. 2). As a control for the developmental changes in NGF mRNA, and to ensure similar amounts of mRNA in each lane, the same blots were reprobed with an α-actin probe. Both the entire embryo and the brain contained similar levels of a- actin mRNA at all stages analysed (Fig. 1). The heart, on the other hand, showed a more complex developmental change in α-actin mRNA levels, with two mRNA species represented at different relative levels at different stages (not shown).
Distribution of NGF mRNA at E8
In order to determine the origin of the high levels of NGF mRNA at E8, a number of organs from embryos of this age was dissected for RNA blot hybridization (Fig. 3). The result clearly showed that NGF mRNA is expressed in many regions of the embryo at E8 (Fig. 3A). Particularly high levels of expression were found in skin (collected mainly from the back of the embryo). The level of expression in skin at E8 exceeds that found in the heart of the adult chicken (Fig. 4). High levels were also found in eye, head (skin, eyes and brain removed) and hind limb (skin removed) (Figs 3, 4). Thigh muscle was collected separately to test if the developing muscles of the limb could account for the high NGF mRNA level found in hind limb. This analysis showed a somewhat lower level of NGF mRNA in the muscle mass than in the entire limb (Fig. 4). The P7 pectoralis muscle contained very low levels of NGF mRNA (not shown).
Brain and spinal cord (Fig. 4), together with spinal ganglia, contained the lowest levels of NGF mRNA of the E8 organs analysed. Viscera (excluding heart but including liver, gizzard, intestine, cloaca, lungs and kidneys) showed a somewhat higher level of NGF mRNA than heart. The yolk sac contained a low but detectable level of NGF mRNA (data not shown). The remaining axial structures of the embryo (excluding skin and limbs but including axial muscles, vertebrae, connective tissue and major nerves and blood vessels) also contained intermediate levels of NGF mRNA (Figs 3,4).
Discussion
The present finding of NGF mRNA as early as E3-5 (stage 20) suggests a function for NGF in early chick embryo development. However, evidence is lacking for the presence of the NGF protein until late in development (day 18; Belew & Ebendal, 1986). Unequivocal demonstration of NGF protein at earlier stages of development would require a sensitive enzyme immunoassay for chicken NGF like that available for rat and mouse NGF (Korsching et al. 1985; Whittemore et al. 1986; Larkfors & Ebendal, 1987; Auburger et al. 1987).
The pattern of NGF mRNA expression in the chicken embryo reveals several interesting features. One obvious characteristic is the presence of NGF mRNA already at a time when the first formation of peripheral ganglia is occurring. The early appearance of NGF mRNA in the chicken embryo at day 3 was recently also observed by Goedert (1986).
The highest level of NGF mRNA expression in early development was found in the skin, an area of sensory innervation, whereas organs densely innervated by adrenergic fibres such as the heart, showed low levels of NGF mRNA before hatch. A peak in the embryonic level of NGF mRNA was recently demon-strated also in skin of the E12·5 mouse embryo by Davies et al. (1987). The distribution of NGF mRNA in the E8 chick embryo, with high levels in the skin and eye, suggest a role for NGF also in exteroceptive sensory innervation during chicken development.
Binding of radiolabelled NGF to receptors has been observed in sections of the E4 chick embryo (Raivich et al. 1985), while premigratory and early migratory neural crest cells during E2 and E3 lack the NGF receptor (Bernd, 1985). The first survival-promoting effects of NGF injected into the chick embryo at days 3 and 4 were seen at E4-5 (stage 25) in the cervical primary sympathetic ganglion (Oppenheim et al. 1982) and thoracic level spinal ganglia (Hamburger et al. 1981).
The peak of NGF mRNA levels (E8-12) coincides with the maximum number of NGF receptors on the spinal chick ganglia (Herrup & Shooter, 1975) and with maximal fibre outgrowth from explanted sympathetic, trigeminal and spinal ganglia responding to mouse NGF (Ebendal & Hedlund, 1975; Ebendal, 1979; Davies & Lindsay, 1984). A similar fibre outgrowth response has recently been demonstrated using a biologically active recombinant chicken NGF protein (Ebendal & Persson, 1987). At around E8, maximal neuronal death in spinal ganglia has also been observed (Hamburger et al. 1981) which has been suggested to be the result of competition between neurones for a limited supply of a trophic factor.
Interestingly, the developing hindlimb contained a high level of NGF mRNA. This high level was not accounted for by the skin (which was removed before preparation of RNA) nor by the developing muscle mass of the thigh. Goedert (1986) also observed substantial amounts of NGF mRNA in the embryonic leg and suggested a functional correlation to the fibre projections from the dorsal root ganglia.
The cellular origin of the NGF mRNA in the skin and leg is not known. Production of NGF may be widespread with several cell types contributing. Moreover, the proportion of cells expressing NGF mRNA at any developmental stage may vary. This issue needs to be resolved by the use of in situ hybridization. Indeed, Davies et al. (1987) recently presented in situ hybridization data from the mouse embryo suggesting synthesis of NGF mRNA in both epithelial and mesenchymal layers of the skin.
The pattern of NGF mRNA expression in the central nervous system is of particular interest. The brain and spinal cord contain relatively high levels of NGF mRNA in the adult chicken (Ebendal et al. 1986; Goedert, 1986). Developmentally, expression of NGF mRNA in these organs increases significantly after hatch with only low levels of NGF mRNA during the embryonic period. A similar time course has been found in the mammalian brain, where NGF mRNA appears postnatally in the rat brain with a peak level three weeks after birth (Whittemore et al. 1986; Large et al. 1986; Auburger et al. 1987). Despite the low levels of NGF mRNA prenatally, the NGF protein was found in the fetal rat brain at levels 25-50% of those found in the adult brain (Whittemore et al. 1986; Large et al. 1986; Auburger et al. 1987). It is not known whether this is also true for the chicken embryo brain. An early function for NGF in brain development is suggested by the presence of the NGF receptor in fetal rat brain (Yan & Johnson, 1987). NGF binding in the chicken brain has been reported at E8 by Frazier et al. (1974) and more recently by Raivich et al. (1987) demonstrating transient, widespread receptor binding of NGF in several areas of the chick brain at E4-12.
A function for NGF in the visual system is also suggested by the finding of NGF mRNA in the embryonic neural retina. Interestingly, optic tectum is the part of adult chicken brain with the highest level of NGF mRNA (Ebendal et al. 1986). Similar observations were made by Goedert (1986) who also showed that the retina maintains a high level of NGF mRNA in adult hens.
The mechanisms regulating the tissue- and stage-specific expression of NGF mRNA revealed by the present study remains to be elucidated.
ACKNOWLEDGEMENTS
Technical assistance was given by Mrs Annika Kylberg, Mrs Stine Söderstrôm, Mrs Britt-Marie Johansson and Mrs Vibeke Nilsson. Support was obtained from the Swedish Medical and Natural Science Research Councils, the Bank of Sweden Tercentenary Foundation and The Swedish Board for Technical Development.